Briffa et al. (1998) describe a reduced sensitivity of recent tree growth to temperature at Northern Hemisphere (NH) high latitudes. Their study shows a divergence starting about 1950 between increasing mean summer temperature and decreasing tree-ring density and ring width for North America and several regions of northern Eurasia. This phenomenon is significant, because it questions the validity of ring-width and density records as being a proxy for mid-to-low frequency temperature variability. We believe that a detailed view of the instrumental records of northern Eurasia will improve our insight in this matter.
Here we analyze the sparse set of meteorological records available for northern Eurasia by subdividing the area North of 65°N into four study regions: North Europe, West Siberia, Central Siberia, and East Siberia. For these regions, we analyzed the raw instrumental data and the adjustments made to these measurements by the Global Historical Climatology Network (GHCN). In particular, we investigated what urban adjustments should be made after the completion of all other station corrections and adjustments. Corrections done for urban warming were of particular interest to us, because rising temperature trends can in part be forced by warming effects of urbanized territories. Our assumption was that the discrepancy between the instrumental and tree-ring data might be caused by corrections that do not adequately or fully remove the urban warming effect.